Vol. 174, No. 1JOURNAL OF BACTERIOLOGY, Jan. 1992, p. 130-1360021-9193/92/010130-07$02.00/0Copyright X 1992, American Society for Microbiology

Escherichia coli purB Gene: Cloning, Nucleotide Sequence,and Regulation by purRt

BIN HE,1 JOHN M. SMITH,2 AND HOWARD ZALKIN1*Department ofBiochemistry, Purdue University, West Lafayette, Indiana 47907,1 and

Seattle Biomedical Research Institute, Seattle, Washington 981092

Received 26 August 1991/Accepted 1 November 1991

Escherichia colipurB encodes adenylosuccinate lyase (ASL), the enzyme that catalyzes step 8 in the pathwayfor de novo synthesis of IMP and also the final reaction in the two-step sequence from IMP to AMP. Gene purBwas cloned and found to encode an ASL protein of 435 amino acids having a calculated molecular weight of49,225. E. coli ASL is homologous to the corresponding enzymes from Bacilus subtilis and chickens and alsoto fumarase from B. subtilis. Gene phoP is 232 bp downstream of purB. Gene purB is regulated threefold bythe purine pool and purR. Transcriptional regulation of purB involves binding of the purine repressor to the16-bp conserved pur regulon operator. The purB operator is 224 bp downstream of the transcription start siteand overlaps codons 62 to 67 in the protein-coding sequence.

Adenylosuccinate lyase (ASL), the product of gene purBin Escherichia coli, catalyzes two reactions in de novopurine nucleotide biosynthesis. In addition to catalyzing step8 in the pathway to IMP, conversion of phosphoribosylsuc-cinocarboxamideaminoimidazole to phosphoribosylamino-imidazole carboxamide, ASL also converts adenylosucci-nate to AMP (Fig. 1). Gene purB has been mapped to 25 minon the E. coli chromosome (2), unlinked to the other purinegenes.The genes for de novo purine nucleotide synthesis in E.

coli, purF, purMN, purEK, purL, purC, purHD, and guaBA,are scattered around the chromosome as monocistronic andsmall polycistronic operons and are negatively controlled bypurR (6, 15). These genes are part of a pur regulon. Nucle-otide sequence analysis ofpur regulon genes has identified aconserved 16-bp sequence generally around the -35 pro-moter region which serves as a site for binding of thepurR-encoded repressor. Gene purB from E. coli has notbeen cloned, and there are conflicting reports concerningregulation by purR. Meng et al. (15) reported 2.5-fold regu-lation of ASL by purR, and He et al. (6) found no regulationby purR of apurB-lacZ fusion made by insertion of Mudl(Aplac).Here we report the nucleotide sequence of cloned E. coli

purB. A binding site for the purine repressor in the protein-coding region is required for threefold regulation by purR.

MATERIALS AND METHODS

Bacterial strains, bacteriophage, and plasmids. The strainsand plasmids used are listed in Table 1. A XSE6 library (4) ofE. coli K-12 DNA fragments was obtained from GrahamWalker, Massachusetts Institute of Technology, Cambridge.Phage XRZ5 has been described (19).

Media. LB (17) and 2XYT (17) were used as rich media.The minimal growth medium used has been described (27).

Cloning of the purB gene. Gene purB was isolated bycomplementation of purB strain TX570 by using a XSE6 E.coli library (4). Phage DNA prepared (24) from a repre-

* Corresponding author.t Journal paper 13209 from the Purdue University Agricultural

Experiment Station.

sentative isolate was partially digested with Sau3AI andelectrophoresed on a 0.5% agarose gel. DNA fragmentsranging in size from 1.5 to 6 kb were excised, electroeluted,and ligated into the BamHI site of pUC118. The ligationmixture was used to transform purB mutant TX530 topurB+.To clone the upstream region of purB, a X phage lysate

was prepared from phage 7F9 of the library of Kohara et al.(8), and DNA was purified (24) as described above. PhageDNA was digested with EcoRI and electrophoresed on a0.8% agarose gel. A 3-kb EcoRI fragment was identified bySouthern blotting (13) by using a 32P-labeled DNA probecorresponding to the 5' end of the purB gene. The 3-kbEcoRI fragment was cloned into pUC118 to give plasmidpBH111.DNA sequence analysis. A series of overlapping exonucle-

ase III deletions was prepared (7), and nucleotide sequenceswere determined by the dideoxynucleotide chain terminationmethod of Sanger et al. (22).Enzyme assay. Cells were grown to the mid-log phase in

minimal medium with and without adenine. Cells weredisrupted (20) in 0.1 M sodium phosphate, pH 7.0, andP-galactosidase activity was determined by the method ofMiller (17). Protein concentration was determined by themethod of Lowry (10).

Repressor-operator binding. For gel retardation assays, 10fmol of a 570-bp HindIII-EcoRI fragment from plasmidpBH113 containing the purB operatorlike sequence wasincubated with the purified Pur repressor in buffer system II(21) with the hypoxanthine corepressor at 50 ,uM and sam-ples were resolved by electrophoresis. For DNase I foot-printing, a DNA fragment was isolated from plasmidpBH113, labeled at one end with [(y-32P]ATP and T4 poly-nucleotide kinase, and then purified by electrophoresis andelectroeluted from a 5% polyacrylamide gel slice. Repressor-operator complexes were prepared in buffer system I (21) byusing 400 ng of the Pur repressor. The conditions used fortreatment with DNase I and for electrophoresis have beendescribed (20).

Primer extension mapping. E. coli R320(pBH111) wasgrown in 50 ml of minimal medium without adenine. At aKlett reading of 70, using a 66 filter, the cells were pouredonto ice chilled to -20°C and collected by centrifugation for

130

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

E. COLI purB GENE 131

SAMP ! AMPPRPP -7 steps *-SAICAR -_-AICAR - -lMP

ASL -_ -0_ GMP

FIG. 1. Outline of the pathway for de novo synthesis of purinenucleotides in E. coli. Abbreviations: PRPP, 5-phosphoribosyl-1-pyrophosphate; SAICAR, 5'-phosphoribosylsuccinocarboxamide-aminoimidazole; AICAR, 5'-phosphoribosylaminoimidazole car-boxamide; SAMP, 6-succinyladenosine-5'-monophosphate.

_ATG

-986 pBH4112 3n

TGA

35 pBH101 1725

FIG. 2. Scheme showing plasmid clones used to determine thepurB nucleotide sequence. The transcription start site (+ 1) isrepresented by the arrow. Numbering is relative to the transcriptionstart site as given in Fig. 3.

4 min at 3,000 x g at 4°C. The cells were then suspended inice-cold TE (10 mM Tris [pH 7.6], 0.1 mM EDTA), and RNAwas isolated by hot-phenol extraction (23). The final RNApellet was dissolved in water and frozen at -70°C.To determine the 5' end of purB mRNA, two 19-mer

oligonucleotide primers, complementary to nucleotides 120to 138 and 154 to 172, were labeled with [y-32P]ATP and T4polynucleotide kinase and annealed to the mRNA. Theconditions used for primer extension mapping with avianmyeloblastosis virus reverse transcriptase have been de-scribed (30).Plasmid constructions. Plasmid pBH113 was constructed

as a purB fragment source for gel retardation assay and

TABLE 1. Strains and plasmids used

E. colistrain Description

MC4100 A(argF-lac)169 Lac-R320 MC4100 purR320BH100 MC4100 (ABLG1) Lac' AprBH101 MC4100 (XBLG2) Lac+ AprBH102 MC4100 (XLT2) Lac+ AprBH103 MC4100 (XLT3) Lac' AprBH200 R320 (XBLG1) Lac' AprBH201 R320 (XBLG2) Lac' AprBH202 R320 (XLT2) Lac' AprBH203 R320 (XLT3) Lac' AprTX570 ara A(lac) purB::Mu AcI857 S7TX530 ara A(lac) I(purB'-lacZ Y+::Xpl

(209, 205)pMLB1034 'lacZ translational fusion vector; AprpRS415 lacZ transcriptional fusion vector; AprpUC118 Phagemid cloning vectorpBH101 1.7-kb Sau3A fragment in BamHI site

of pUC118; purB+pBH111 3-kb EcoRI fragment from purB+ phage

7F9 in EcoRI site of pUC118pBH112 1.3-kb KpnI-EcoRI fragment from

pBH111 in pUC118pBH113 570-bp BglI-EcoRI fragment from

pBH111 in pUC118pBLG1 1.1-kb SmaI-BamHI fragment

(nucleotides -986 to 172 in SmaI-BamHI sites of pMLB1034

pBLG2 1.3-kb SmaI-BamHI fragment(nucleotides -986 to 316) inpMLB1034

pLT1 0.8-kb Smal-BamHI fragment(nucleotides -986 to -172) in SmaI-BamHI sites of PRS415

pLT2 1.1-kb SmaI-BamHI fragment(nucleotides -986 to 172) in SmaI-BamHI sites of PRS415

pLT3 1.3-kb SmaI-BamHI fragment(nucleotides -986 to 316) in PRS415

pLT4 350-bp Smal-BamHI fragment(nucleotides -172 to 172) in pRS415

Source orreference

1919This workThis workThis workThis workThis workThis workThis workThis workThis work28

242526This work

This work

This work

This work

This work

This work

This work

DNase I footprinting. First, plasmid pBH111 was digestedwith BglI and treated with T4 DNA polymerase to make ablunt end. Following this step, the plasmid was digested withEcoRI. A 570-bp BglI-EcoRI fragment containing the purBoperatorlike sequence was subcloned into the HincII andEcoRI sites of pUC118 to yield pBH113. Plasmid pBH112was constructed by subcloning a 1.3-kb KpnI-EcoRI purBfragment from pBH111 into pUC118 (Table 1).For construction of translational and transcriptional purB-

lacZ fusions, purB DNA was obtained from plasmid pBH112by polymerase chain reaction amplification (18). For trans-lational fusions, the 5' end of purB was at nucleotide -986and contained an SmaI site. The 3' end with a BamHI sitewas at nucleotide 172 for pBLG1 (purBo-) or 316 for pBLG2(purBo+). Fragments were ligated into plasmid pMLB1034to give purB-lacZ translational fusions pBLG1 (purBo-) andpBLG2 (purBo+). In pBLG1 (purBo-) and pBLG2(purBo+), purB codons 45 and 93, respectively, were fusedto codon 8 of lacZ. For transcriptional fusions, fragmentswere ligated into the SmaI-BamfHI sites of pRS415. PlasmidpLT1 (purBo-) contains purB nucleotides -986 to -172,pLT2 (purBo-) contains nucleotides -986 to 172, pLT3(purBo+) contains nucleotides -986 to 316, and pLT4(purBo-) contains nucleotides -172 to 172. purB-lacZ tran-scriptional and translational fusions were recombined intoXRZ5 for insertion into the chromosome in single-copy form(19).

N-terminal ASL amino acid sequence. An ASL.--galac-tosidase chimeric enzyme was purified from E. coliR320(pBLG1). The enzyme was isolated from a cell extractby ammonium sulfate precipitation (0 to 45% saturation)followed by immunoaffinity chromatography on an anti-,B-galactosidase agarose column (Promega). The eluted enzymewas dialyzed against 10 mM NH4HCO3 and then lyophilized.A 50-,ug sample of protein was electrophoresed on an 8%sodium dodecyl sulfate-polyacrylamide gel, and proteinswere transferred to a polyvinylidene difluoride-type mem-brane and lightly stained with Coomassie blue (12). Theprotein band was cut out and subjected to amino acidsequencing by the Purdue Laboratory for MacromolecularStructure.

Nucleotide sequence accession number. The sequence ofgene purB has been deposited in the GenBank data baseunder accession no. M74924.

RESULTS

Cloning of purB. Gene purB was isolated from a library ofE. coli DNA fragments by complementation of E. coli purB

This work mutant TX570 as described in Materials and Methods. Onerepresentative plasmid, designated pBH101, containing aninsert of 1.7 kb was found to contain the intact purB coding

This work sequence but to lack the 5' flanking region (Fig. 2). On theThis work basis of this information, the purB upstream region was

isolated on a 3-kb EcoRI fragment from X phage 7F9 from thelibrary of Kohara et al. (8). This 3-kb EcoRI fragment was

VOL. 174, 1992

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

132 HE ET AL.

GGTACCGAAGAACCGTGGTATGTGGTGGACAAAGACGTCGAAAACAACATTCTGGTTGTC

GCTCAGGGCCATGAACACCCGCGGCTGATGTCTGTCGGGTTGATTGCCCAGCAGTTGCAC

CAGACCGATATCCCTTGCACCGTCAAGGCGCTGGACGATGATCGCATTGAAGTGATTTTC

GATGAACCGGTTGCCGCCGTGACGCCGGGCCAGTCTGCCGTCTTCTATAACGGTGAAGTG

CTTAAAACAAGGAAGCAGTGAACGTGGCAAAGAATTACTATGACATCACCCTCGCCCTGG

CCGGTATTTGTCAGTCGGCACGCCTGGTGCAACAACTCGCTCACCAGGGGCATTGTGATG

CCGATGCGCTACACGTCTCACTCAACAGTATTATTGATATGAACCCAGCTCGACGCTGGC

CAATGCCAGCAGTCGCCAGGGCTTAAACGCCGAATTAACCCGCTACACACTCAGCTTGAT

GGTGCTTGAGCGCAAACTCTCCTCAGCGAAAGGCGCGCTCGACACTCTGGGCAACCGGAT

GATGGCTGCTATCTATGTTGATGTGATTAGCCCGCTTGGCCCGCGCATTCAGGTCACCGG

TTCCCCTGCTGTACTGCAAAGCCCACAAGTGCAGGCGAAAGTTCGCGCAACCCTGCTGGC

AGCATTCGCGCGCCGTGCTCTGGCACCAGGTCGGCGGCGGACGTCTGCAACTGATGTTTT+1 srx

CTCGTAATCGCCTGACCACTCAGGCAAAACAAATTCTTGCTCATTTAACCCCGGAGTTGTpurB

GATCTATGGAATTATCCTCACTGACCGCCGTTTCCCCTGTCGATGGACGCTACGGCGATAMetGluLeuSerSerLeuThrAlaValSerProValAspGlyArgTyrGlyAspL

AAGTCAGCGCGCTGCGCGGGATTTTCAGCGAATATGGTTTGCTGAAATTCCGTGTACAAGysValSerAlaLeuArgGlyIlePheSerGluTyrGlyLeuLeuLysPheArgValGlnV

TTGAAGTACGTTGGCTGCAAAAACTGGCCGCGCACGCAGCGATCAAGGAAGTTCCTGCTTalGluValArgTrpLeuGlnLysLeuAlaAlaHisAlaAlaIleLysGluValProAlaP

TTGCTGCCGCGCAATCGGTTACCITGATGCAATCGTCGCCAGTTTCAGCGAAGAAGATGheAlaAlaAspAlaIleGlyTyrLeuAspAlaIleValAlaSerPheSerGluGluAspA

TTGCCTGTACTTCGGAAGATATCAATAACCTCTCCCACGCATTAATGCTGAAAACCGCGCheAlaCysThrSerGluAspIleAsnAsnLeuSerHisAlaLeuMetLeuLysThrAlaA

rgAspGluValIleLeuAlaTyrTrpArgGlnLeulleAspGlyLeuLysAspLeuAlaV

TTCAGTATCGCGATATCCCGCTGCTGTCTCGTACCCACGGTCAGCCAGCCACGCCGTCAAalGlnTyrArgAspIleProLeuLeuSerArgThrlisGlyGlnProAlaThrProSerT

CCATCGGTAAAGAGATGGCAAACGTCGCCTACCGTATGGAGCGCCAGTACCGCCAGCTTAhrIleGlyLysGluMetAlaAsnValAlaTyrArgMetGluArgGlnTyrArgGlnLeuA

CCGCTTACCCGGAAGTTGACTGGCATCAGTTCAGCGAAGAGTTCGTCACCTCGCTGGGTAlaAlaTyrProGluValAspTrpHisGlnPheSerGluGluPheValThrSerLeuGlyI

TTCAGTGGAACCCGTACACCACCCAGATCGAACCGCACGACTACATTGCCGAACTGTTTGleGlnTrpAsnProTyrThrThrGlnIleGluProHisAspTyrIleAlaGluLeuPheA

ATTCGGTTGCGCGCTTCAACACTATTCTGATCGACTTTGACCGTGACGTCTGGGGTTATAspSerValAlaArgPheAsnThrIleLeuIleAspPheAspArgAspValTrpGlyTyrI

TCGCCCTTAACCACTTCAAACAGAAAACCATTGCTGGTGAGATTGGTTCTTCCACCATGCleAlaLeuAsnHisPheLysGlnLysThrIleAlaGlyGluIleGlySerSerThrMetP

CGGTATTGCAGACTCTGGCAAGCAAACTGCCGGTTTCCCGCTGGCAGCGTGACCTGACCGlaValLeuGlnThrLeuAlaSerLysLeuProValSerArgTrpGlnArgAspLeuThrA

ACTCTACCGTGCTGCGTAACCTCGGCGTGGGTATCGGTTATGCCTTGATTGCATATCAATspSerThrValLeuArgAsnLeuGlyValGlyIleGlyTyrAlaLeuIleAlaTyrGlnS

CCACCCTGAAAGGCGTGAGCAAACTGGAAGTGAACCGTGACCATCTGCTGGATGAACTGG

ATCAAACTGGGAAGTGCTGGlLCTAACCAATCClAGACAGTTtGACrGTCArTGGCATlGspHisAsnTrpGluValLeuAlaGluProI leGInThrValtietArgArgTyrGlyI leG

TGAAGCAGTTTATCGATGGTCTGCGTTGCCAGAAGAAGAGAAAGCCCGCCTGAAAGCGATetLysGlnPheIleAspGlyLeuArgCysGlnLysLysArgLysProAlaEnd

TTCACTTTACCTCCCCTCCCCGCTGGTTTATTTAATGTTTACCCCCATAACCACATAATCJsl phoP

GCGTTACACTATTTTAATAATTAAGACAGGGAGAAATAAAAATGCGCGTACTGGTTGTTGMetArgValLeuValValG

927 initially subcloned into the EcoRI site of pUC118, giving-867

pBH111. Next, a 1.3-kb KpnI-EcoRI subfragment from-807 pBH111 was transferred to pUC118 to yield pBH112. The-807 1.3 kb of E. coli DNA in pBH112 overlapped with pBH101

and extended the purB 5' flanking region (Fig. 2).-687 Nucleotide sequence of purB. The nucleotide sequence of-627 purB and the 5' and 3' flanking regions was determined on-567 both DNA strands from plasmids pBH101 and pBH112 (Fig.-507 3). In purB+ plasmid pBH101, there is a single open reading_447 frame of 1,305 bp that encodes a protein of 435 amino acids-387 with an Mr of 49,225. This open reading frame is ascribed to

-327ASL on the basis of complementation of purB strain TX530-327 and homology with Bacillus subtilis (3) and avian (1) ASL.

-267 The NH2-terminal amino acid sequence of ASL from a

-207 purB-lacZ translational fusion in plasmid pBLG1 was deter-

-147 mined and found to correspond exactly to residues 1 to 6 of-87 the amino acid sequence shown in Fig. 3.-27 A 16-bp sequence with an imperfect dyad symmetry at34 nucleotides 224 to 239 in the protein-coding region has

94 identity with the pur regulon operator in 11 of 16 positions19 (Fig. 4). All highly conserved bases are found in the purB

154 operatorlike sequence, except for the A at left-hand position2 and the T at right-hand position 2. Evidence that this pur

214

59 operatorlike sequence is involved in purR control is pre-274 sented below.79 Approximately 230 nucleotides downstream ofpurB, there

334 is an open reading frame beginning with an ATG at nucleo-99 tide 1577. This open reading frame is preceded by a good

13949 Shine-Dalgarno sequence and extends for 50 codons to the

454 end of the fragment. A search of the protein data base with139 FASTA (11) indicated 90% identity with Salmonella typhi-514 murium phoP over the first 50 amino acids. We did not159

574 investigate phoP expression. However, there was close179 correspondence of nucleotides 1344 to 1349 and 1368 to 1373634 to the -35 and -10 promoter elements, respectively.199 Upstream of purB, at nucleotides -391 to -23, there is an

61949 open reading frame with the potential to encode a 123-amino-

754 acid protein chain. This open reading frame is not preceded239 by a Shine-Dalgarno sequence, and translation is considered814 unlikely. No significant similarities to this sequence were259 found in release 28 of the National Biomedical Research874 Foundation protein data base.279

934 purB promoter. Three transcriptional fusion plasmids were299 constructed to detect purB promoter activity. Fragments of994 DNA corresponding to nucleotides -986 to -172, -172 to319 172, and -986 to 172 were inserted into the SmaI-BamHI

1054 sites of plasmid pRS415. This plasmid contains an intact339

1114 lacZ coding sequence and ribosome-binding site but lacks a359 promoter. P-Galactosidase activity was measured from con-

1174 structs transformed into strain MC4100. The data in Fig. 5379 indicate that both subfragments of the upstream region,

1234 nucleotides -986 to -172 and -172 to 172, have the399

1294 capacity for purB promoter function. Plasmid pLT2, con-419 taining E. coli DNA nucleotides -986 to 172, had 1.7- to

1354 2.0-fold more activity than either subfragment.435 Primer extension mapping was performed to identify purB

17414 mRNA 5' ends. With a primer complementary to nucleotides

1534 154 to 172, a single major cDNA was obtained ending at

15957

FIG. 3. Nucleotide sequence ofpurB and the derived ASL aminoacid sequence. Numbering is from the experimentally determinedstart of transcription. The 16-bp conserved pur regulon operator isboxed. The sequence of phoP follows that of purB. Shine-Dalgarno(SD) sites are overlined.

J. BACTERIOL.

AAGACAATGCGTTGTTACGTCACCACCTTAAAGTTCAGATTCAGGATGCTGGTCATCAGG 1654luAspAsnAlaLeuLeuArgHisHisLeuLysValGlnIleGlnAspAlaGlyHisGlnV 27

TCGATGACGCAGAAGATGCCAAAGAAGCCGATTATTATCTCAATGAACATATACCGGATA 1714alAspAspAlaGluAspAlaLysGluAlaAspTyrTyrLeuAsnGluHisIleProAspI 47

TTGCGATTGTC 1725leAlaIleVal 50

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

E. COLI purB GENE 133

8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8

purF A E E E E EEE E- I I I TE T IpurMNT ElE MENpurL A -* G |purC A EE| * * - i G

G A C T

G E T

|| T I G lII G N G i

purEK A C N - EI I I T C §purHD G i i - I I I I T G I

purRj A G E| * I9 I A C I

purR2 G

purB A

A EN IE - GI NC AC

M MEI T MGINoA EC I

FIG. 4. Alignment of the purB operator with other pur regulonoperators. Consensus positions that are conserved in six or moreoperators are highlighted.

adenine +1 (Fig. 6). The same mRNA 5' end was mappedwith a second primer complementary to nucleotides 120 to138 (data not shown). In a separate experiment, the samemRNA 5' end was mapped by using RNA from cells carryingpurB-lacZ plasmid pBLG1 (data not shown). The transcrip-tion start site is thus designated as adenine + 1. This positionis not preceded by a typical a70 promoter sequence. Addi-tional upstream mRNA 5' ends were not obtained by primerextension mapping with an oligonucleotide corresponding tonucleotides -164 to -147. Therefore, the promoter for purBtranscription in the upstream region between nucleotides-986 to -172 in plasmid pLT1 remains to be identified.Regulation by purR. A series of experiments was con-

ducted to determine whether purB is subject to regulation bypurR and whether the pur operatorlike sequence in the purBcoding region has a regulatory function. Transcriptional andtranslational fusions to lacZ were constructed to investigatein vivo regulation by purines and purR. DNA fragmentscontaining nucleotides -986 to 172 and -986 to 316 (Fig. 3)were ligated into vectors to give transcriptional lacZ fusionplasmids pLT2 and pLT3 and translational lacZ fusions inplasmids pBLG1 and pBLG2 (Fig. 7). Plasmids pLT2 andpBLG1 lack the operatorlike sequence, whereas pLT3 andpBLG2 contain the putative purB operator. These purB-lacZfusions were integrated into the chromosome for assay of,-galactosidase. The data summarized in Table 2 demon-

-986

pLT1 E-172 -alactosidase

823

+1i ATG 172-986

pLT2 [

-172pLT4

+1I ATG 172

699

GCTA

TATA

GCTATA

tA

FIG. 6. Primer extension mapping of the 5' end of purB mRNA.RNA was isolated from strain R320(pBH111). The same primer wasused for synthesis of cDNA (lanes 1 and 2) and the sequencingladder (lanes G, A, C, and T). The nucleotide sequence and mRNA5' end corresponding to the site for transcription initiation are shownon the right.

strate that with purB-lacZ transcriptional fusions in strainsBH102, BH202, BH103, and BH203 and with translationalfusions in strains BH100, BH200, BH101, and BH201,repression by adenine was dependent upon purR+ and thepurB operatorlike site. The cis-acting purB control site is

pLT2/pBLG1-986 rF- ATG 172

1 45

pLT3/pBLG2-986

pursor L- ATG 316

FIG. 5. Detection ofpurB promoter function. The boxes providea schematic representation ofpurB 5' flanking DNA cloned into lacZtranscriptional fusion vector pRS415. The shaded segment corre-sponds to the protein-coding region. Numbers refer to nucleotidepositions. P-Galactosidase activity (nanomoles per minute per mil-ligram of protein) is given for MC4100 transformants. The 3-galac-tosidase control with the vector alone gave 35 U of P-galactosidaseactivity.

1 93FIG. 7. Schematic representation of purB 5' flanking DNA used

for construction of transcription and translational lacZ fusions.Transcriptional fusions in pRS415 are pLT2 and pLT3. Translationalfusions in pMLB1034 are pBLG1 and pBLG2. Numbers above thediagrams refer to the nucleotide sequence, and those below refer tothe amino acid sequence. The horizontal arrows define the transcrip-tion start sites. The purB operator is designated purB,.

1 2

VOL. 174, 1992

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

134 HE ET AL.

TABLE 2. Regulation of purB-lacZ

Sr -Galactosidasea sp act FoldStrain purR rersinb+Adenine -Adenine repression

BH102 (purBo-) + 151 158 1.0BH202 (purBo-) - 150 144 1.0BH103 (purBo+) + 46.7 152 3.1-3.3BH203 (purBo+) - 143 146 1.0BH100 (purBo-) + 63.3 56.4 0.9-1.0BH200 (purBo-) - 60.0 57.3 1.0BH101 (purBo+) + 17.8 61.1 3.3-3.4BH201 (purBo+) - 58.3 59.2 1.0

a -Galactosidase values are expressed in specific activity as averages ofthree or four experiments with a variation of 1Ot% or less.

b Fold repression was calculated two ways and expressed as a range:enzyme activity with adenine divided by that without adenine in purR+ andenzyme activity with adenine in purR divided by that in purR+.

thus confirmed to function as an operator (purBo+). Repres-sion of purB-lacZ by adenine was 3.1- to 3.4-fold in thetranscriptional and translational fusions. There was no purBrepression by added purine in purR or purBo- strains,indicative of a single repressor-operator system for control.Therefore, transcriptional repression by adenine requiresnot only purR+ but also purB DNA containing the operator-like sequence.

Binding of the Pur repressor to purBo. Since in vivoregulation of purB requires a segment of the purB codingsequence containing the 16-bp cis-acting control site, it wasimportant to demonstrate that the Pur repressor binds to thissequence. Accordingly, gel retardation and DNase I foot-printing experiments were performed to evaluate specificbinding. The results of a typical gel retardation experimentare shown in Fig. 8. This experiment shows binding of thepurine repressor to a 570-bp purB DNA fragment corre-sponding to nucleotides 221 to 242. Approximately 80 ng ofthe repressor was required for 50% binding. Binding of therepressor was dependent upon the purine corepressor andwas specific forpurB DNA containing the 16-bp pur operator(data not shown).To define the site of repressor-DNA interaction, DNase I

1 2 3 4 5 6 7

A 1 234 A GCT

B221 244l_ . lCCGI ACGCMTCGGTTACCTITGATG

FIG. 9. DNase I footprint of the purB-repressor complex. (A)Lanes: 1 and 4, 570-bp HindIII-EcoRI DNA from pBH113 labeled atthe EcoRI site; 2 and 3, the same DNA plus 0.4 ,ug of the purinerepressor; A, G, C, and T, sequencing ladder from pBH113. (B)Nucleotide sequence of the region protected by the repressor. The16-bp purB operator is boxed.

FIG. 8. Gel retardation assay for binding of the purine repressorto purBo. All lanes contained 10 fmol of the 570-bp HindIII-EcoRIDNA fragment from pBH113. Lanes 1 to 7 contained 0, 5, 10, 20, 40,80, and 160 ng of the purine repressor. Free DNA and DNA-proteincomplexes were resolved on a 4% polyacrylamide gel.

footprinting was conducted. The results of a typical experi-ment are presented in Fig. 9, which shows that the Purrepressor bound specifically to the 16-bp pur operator in the570-bp DNA fragment that was retarded in the experimentshown in Fig. 8. No other positions in the DNA fragmentwere protected against DNase I cleavage. This confirms thatthe operatorlike sequence in the purB coding region func-tions as a site for repressor binding and this repressor-DNAinteraction is required for purB regulation.

DISCUSSION

E. coli purB was cloned by functional complementation ofa purB mutant. The DNA sequence predicts an ASL proteinof 435 amino acids having a molecular weight of 49,225. E.

J. BACTERIOL.

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

E. COLI purB GENE 135

TABLE 3. Summary of E. coli ASL alignments'

Enzyme EcASL BsASL GgASL EcFUM BsFUM EcASP HuRSL

EcASL 100BsASL 26 (29) 100GgASL 22 (15) 26 100EcFUM 19 (6.0) 22 24 100BsFUM 21 (12) 22 21 63 100EcASP 20 (6.5) 21 20 27 32 100HuRSL 19 (3.4) 21 22 22 23 20 100

a The numbers are percentages of identical amino acids for pairwisecomparisons of E. coli ASL (EcASL), B. subtilis ASL (BsASL) (3), chickenASL (GgASL) (1), E. coli fumarase (EcFUM) (29), B. subtilis fumarase(BsFUM) (16), human argininosuccinate lyase (HuRSL) (14), and E. coliaspartase (EcASP) (29). The numbers in parentheses are Z values (standarddeviations above the alignment of randomized sequences).

coli ASL was aligned with ASL sequences from B. subtilisand chickens, as well as sequences of related enzymes thatcatalyze p-elimination reactions. These results are summa-rized in Table 3. There was 26% identity with the enzymefrom B. subtilis and 22% identity with avian ASL. Bothcomparisons had Z values of >10 standard deviations aboverandom, indicating a highly significant probability of relat-edness. Interestingly, the data in Table 3 indicate significantsimilarity between ASL and B. subtilis fumarase. The Zvalue of 12 suggests an evolutionary relationship for theseenzymes. E. coli ASL, E. coli fumarase, and human argini-nosuccinate lyase have several blocks of identical aminoacids., although the Z values are insufficient to supporthomology. The most highly conserved region in these en-zymes is between residues 294 and 308, on the basis of E.coli ASL numbering (Fig. 10). This block of residues hasbeen proposed to represent a "signature sequence" forenzymes that catalyze ,-elimination reactions that generatefumarate as a product (1). The previously suggested mech-anistic similarity among fumarase, ASL, and argininosucci-nate lyase (5) thus reflects a similarity in primary structureand a likely evolutionary relationship.The data summarized in Table 2 indicate that purB expres-

sion is regulated by the purine pool and purR. Guanine andhypoxanthine are corepressors, needed for Pur repressorbinding to the DNA control site (21). With single-copypurB-lacZ chromosomal integrants, there was threefold reg-ulation by purR. This transcriptional regulation requires the16-bp purB operator sequence located 224 nucleotides down-stream from the transcription start site in the protein-codingregion. Whereas this location for the cis-acting control siterelative to the promoter is different from that in the other sixoperons involved in IMP synthesis, it resembles that inpurR. For the other structural genes required for IMPsynthesis, the pur operator is located between positions -46

HuRSL - - -G S S LM P Q K K N P D S L E - - - -

BsFUM - - -G S S I MP G K V N P T Q 9 E - - - -

EcFUM - G S S I 2 P A K V N P V V P E - - - -

EcASP - - - G S S I k P A K V N P V V P E - - - -

GgASL - - -G S S A X P Y K R N P H R S E - - - -

BsASL - - - G S S A 1 P H K R N P I G S E - - - -

EcASL - - - G S S T N P H K V N P I D F E

FIG. 10. Sequence alignment of a conserved region implicated asa signature sequence for ,-elimination reactions that generate fuma-rate. Sequences are ASL from E. coli (Ec), B. subtilis (Bs) (3), andchickens (Gg) (1); fumarase (FUM) from E. coli (29) and B. subtilis(16); and argininosuccinate lyase (RSL) from humans (Hu) (14).

and + 10 relative to the start of transcription (6). However, inpurR there are dual operators at positions 96 to 111 and 184to 199 relative to the start of transcription (20). The down-stream operator, purRo2, overlaps codons 10 to 15 in purR.In purR, operators 01 and 02 contribute 2.5- to 2.7-foldtranscriptional repression, whereas purRo2 by itself contrib-utes only about 1.5-fold regulation by the repressor. Thusrepressor control at purBo is about twofold more effectivethan the regulation at purRo2, even though purBo is 41nucleotides farther downstream from the site for transcrip-tion initiation and is situated deeper in the protein-codingregion that overlaps codons 62 to 67. As shown in Fig. 4,there are only minor nucleotide sequence differences be-tween purBo and purRo2. There is one further strong simi-larity between purB and purR. Neither gene has a goodconsensus -35 and -10 u70 promoter sequence next to thesite mapped for transcription initiation.The regulation of the other pur genes is over a 5- to 17-fold

range (6) and is likely determined by the competing rates ofrepressor-operator and RNA polymerase-promoter complexformation. Large differences in lac operon repression havebeen found to result from placement of the operator indifferent positions relative to the promoter, and it wasconcluded that the mechanism of repression of transcriptioninitiation is dependent upon relative operator-promoter po-sitioning (9). The mechanism by which a repressor that bindsto a DNA site greater than 200 bp downstream from thepromoter can regulate transcription is not known. It wouldappear to be necessary for the bound repressor either tointerfere with transcription initiation or to act as a road blockand obstruct the progress of the transcribing polymerase.The threefold regulation of purB by purR confirms the

earlier observation by Meng et al. for repression of ASL bypurR (15). Previously, we (6) found no regulation of 4(purB-lacZ)205 made in vivo by Mudl(Ap lac) insertion (28),although two- to threefold regulation had been noted in theoriginal report (28). The position of the purB fusion to lacZin 4(purB-lacZ)205 is not known, and therefore we do notknow whether purBo is included in the fusion.

ACKNOWLEDGMENTS

We thank our colleagues Ronald Somerville and Barry Wanner forproviding phage 7F9 and plasmid pRS415, respectively, Kang YellChoi for providing the pure PurR repressor, and Gaochao Zhou forimportant suggestions and discussion.

This work was supported by Public Health Service grantsGM24658 (H.Z.) and A120068 (J.M.S.) from the National Institutesof Health. Computer facilities were supported by NIH grantAI27713.

REFERENCES1. Aimi, J., J. Badylak, J. Williams, Z. D. Chen, H. Zalkin, and

J. E. Dixon. 1990. Cloning of a cDNA encoding adenylosucci-nate lyase by functional complementation in Escherichia coli. J.Biol. Chem. 265:9011-9014.

2. Bachmann, B. 1983. Linkage map of Escherichia coli K-12,edition 7. Microbiol. Rev. 47:180-230.

3. Ebbole, D. J., and H. Zalkin. 1987. Cloning and characterizationof a 12-gene cluster from Bacillus subtilis encoding nine en-zymes for de novo purine nucleotide synthesis. J. Biol. Chem.262:8274-8287.

4. Elledge, S. J., and G. C. Walker. 1985. Phasmid vectors foridentification of genes by complementation of Escherichia colimutants. J. Bacteriol. 162:777-783.

5. Havir, E., and K. Hanson. 1973. The enzymic elimination ofammonia, p. 75-166. In P. D. Boyer (ed.), The enzymes, 3rded., vol. 7. Academic Press, Inc., New York.

6. He, B., A. Shiau, K. Y. Choi, H. Zalkin, and J. M. Smith. 1990.

VOL. 174, 1992

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

136 HE ET AL.

Genes of the Escherichia coli pur regulon are negatively con-trolled by a repressor-operator interaction. J. Bacteriol. 172:4555-4562.

7. Henikoff, S. 1987. Unidirectional digestion with exonuclease IIIin DNA sequence analysis. Methods Enzymol. 155:156-165.

8. Kohara, Y., K. Akiyama, and K. Isono. 1987. The physical mapof the whole E. coli chromosome: application of a new strategyfor rapid analysis and sorting of a large genomic library. Cell50:495-508.

9. Lanzer, M., and H. Buj4rd. 1988. Promoters largely determinethe efficiency of repressor action. Proc. NatI. Acad. Sci. USA85:8973-8977.

10. Layne, E. 1957. Spectrophotometric and turbidometric methodsfor measuring proteins. Methods Enzymol. 3:448-451.

11. Lipman, D. J., and W. R. Pearson. 1988. Improved tools forbiological sequence comparison. Proc. Natl. Acad. Sci. USA85:2444-2448.

12. Madsudaira, P. 1987. Sequence from picomole quantities ofproteins electroblotted onto polyvinylidene difluoride mem-brane. J. Biol. Chem. 262:10035-10038.

13. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual, p. 383-389. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

14. Matuo, S., M. Tatsuno, K. Kobayashi, T. Saheki, T. Miyata, S.Iwanaga, Y Amaya, and M. Mori. 1988. Isolation of cDNAclones of human argininosuccinate lyase and corrected aminoacid sequence. FEBS Lett. 234:395-399.

15. Meng, L. M., M. Kilstrup, and P. Nygaard. 1990. Autoregula-tion ofpurR repressor synthesis and involvement ofpurR in theregulation of purB, purC, purL, purMN and guaBA expressionin Escherichia coli. Eur. J. Biochem. 187:373-379.

16. Miles, J. S., and J. R. Guest. 1985. Complete nucleotidesequence of the fumarase gene (citG) of Bacillus subtilis 168.Nucleic Acids Res. 13:131-140.

17. Miller, J. H. 1972. Experiments in molecular genetics, p.403-404. Cold Spring Harbor Laboratory, Cold Spring Harbor,N.Y.

18. Mullis, K. B., and F. A. Faloona. 1987. Specific synthesis ofDNA in vitro via a polymerase-catalyzed chain reaction. Meth-

ods Enzymol. 155:335-350.19. Rolfes, R. J., and H. Zalkin. 1988. Regulation ofEscherichia coli

purF. Mutations that define the promoter, operator, and purinerepressor gene. J. Biol. Chem. 263:19649-19652.

20. Rolfes, R. J., and H. Zalkin. 1988. Escherichia coli gene purRencoding a repressor protein for purine nucleotide synthesis.Cloning, nucleotide sequence and interaction with the purFoperator. J. Biol. Chem. 263:19653-19661.

21. Rolfes, R. J., and H. ZaIkin. 1990. Purification of the Esche-richia coli purine regulon repressor and identification of core-pressors. J. Bacteriol. 172:5637-5642.

22. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

23. Shimotsu, H., M. I. Kuroda, C. Yanofsky, and D. J. Henner.1986. Novel form of transcriptional attenuation regulatesexpression of the Bacillus subtilis tryptophan operon. J. Bacte-rioL. 166:461-471.

24. Silhavy, T. J., M. L. Berman, and L. W. Enquist. 1984.Experiments with gene fusions, p. 95-96. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

25. Simons, R. W., F. Houmap, and N. Klechner. 1987. Improvedsingle and multicopy Lac-based cloning vector for protein andoperon fusions. Gene 53:85-96.

26. Vieira,-J., and J. Messing. 1982. The pUC plasmids, anM13mp7-derived system for insertion mutagenesis and sequenc-ing with synthetic universal primers. Gene 19:259-268.

27. Vogel, H. T., and D. M. Bonner. 1956. Acetylornithinase ofEscherichia coli: partial purification and some properties. J.Biol. Chem. 218:97-106.

28. Wolfe, S. A., and J. M. Smith. 1985. Separate regulation ofpurAand purB loci of Escherichia coli K-12. J. Bacteriol. 162:822-825.

29. Woods, S. A., J. S. Miles, R. E. Roberts, and J. R. Guest. 1986.Structural and functional relationships between fumarase andaspartase. Biochem. J. 237:547-557.

30. Zheng, L. M., P. C. Andrews, M. A. Hermodson, J. E. Dixon,and H. Zalkin. 1990. Cloning and structural characterization ofporcine heart aconitase. J. Biol. Chem. 265:2814-2821.

J. BACTERIOL.

on May 24, 2020 by guest

http://jb.asm.org/

Dow

nloaded from